On Fri, 12 Mar 2010, Tim Bunce wrote:
> On Fri, Mar 12, 2010 at 04:33:24PM +0100, Eissfeldt, Heiko wrote:
> >    Could you please fix the original XSUB.h also, when the perl5
> >    code freeze is over?
> >
> >    BTW: What do you think about a complete set of tests for all file
> >    I/O function mappings from the API with all different I/O-related
> >    #defines combinations?
> >
> >    We could setup a dummy xs file with all file I/O functions used
> >    and run it through the C compiler to catch any parameter type
> >    mismatches...
> >
> >    That way we could catch these errors once and get rid of them for
> >    ever :-)
>
> [I presume you're talking about Perl and to Jan.]

It is not clear to me how this would catch the errors automatically,
as there may not be parameter type mismatches.  That has been the
problem with swapped parameters for fputs() and fputc(), but that is
different from the missing redefine for fgets().

You also cannot test all different sets of #defines without building
Perl with the corresponding #defines as well, so even if you could
automate the "catching" part, the comprehensive testing would still
require an additional harness around that.

But feel free and go ahead if you think you could pull it off!  The
more people trying to rid Perl of bugs the merrier! :)

> Jan, is there a corresponding perl RT ticket for the fgets
> redefinition?

No, there isn't (AFAIK).  I was just planning to fix this in my
jdb/bug72704 topic branch and maybe push it to the central repo
instead of just github.  That would be enough of a reminder for
me. :)  But feel free to create an additional bug for this (or
rather the errno case, as the fgets() fix is just a single line).

Cheers,
-Jan


-- 
You've received this message because you are subscribed to
the Devel::NYTProf Development User group.

Group hosted at:  http://groups.google.com/group/develnytprof-dev
Project hosted at:  http://perl-devel-nytprof.googlecode.com
CPAN distribution:  http://search.cpan.org/dist/Devel-NYTProf

To post, email:  [email protected]
To unsubscribe, email:  [email protected]

Reply via email to